Non‐vitamin K antagonist oral anticoagulants (NOACs) post‐percutaneous coronary intervention: a network meta‐analysis

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To review the evidence from randomised controlled trials assessing the efficacy of non‐vitamin K antagonist oral anticoagulants (NOACs) post‐percutaneous coronary intervention (PCI) in people with an indication for anticoagulation.

Background

Description of the condition

Coronary artery disease (CAD) is the leading cause of death worldwide (Benjamin 2018; Finegold 2013). According to the World Health Organization (WHO), CAD is responsible for about 15% of all deaths globally (WHO 2015). It remains one of the world's largest health problems despite dramatic medical advances over the last few decades.

CAD is usually caused by atherosclerosis, where fatty deposits accumulate on the walls of the coronary arteries. In general, CAD is classified into stable coronary artery disease (SCAD) and acute coronary syndrome (ACS), the latter of which has three forms: unstable angina (UA), non‐ST‐elevation myocardial infarction (NSTEMI), and ST‐elevation myocardial infarction (STEMI) (Ibanez 2018; Roffi 2016; Wong 2014).

Percutaneous coronary intervention (PCI) is the most common procedure used in the invasive treatment of people with CAD (Khera 2016), and accounts for 3.3% of all operating room procedures performed in USA in 2014 (McDermott 2017). Every year, millions of patients undergo PCI, which is a non‐surgical revascularisation technique used to widen narrowed or blocked blood coronary arteries in order to restore or improve blood flow to the heart muscle (McGrath 1999; Peterson 2000). PCI is indicated in ACS, and may also be performed in people with SCAD who are dissatisfied with their quality of life, particularly if their symptoms are uncontrolled with medication (Khera 2016; Windecker 2014).

Antithrombotic therapy is required after PCI to reduce the risk of recurrent cardiovascular events and stent thrombosis (Leon 1998; Valgimigli 2017). Long‐term dual antiplatelet therapy (DAPT) with aspirin and a P2Y12 receptor antagonist has been shown to reduce mortality and morbidity post‐PCI (Atar 2014; Mauri 2014; Wallentin 2009; Wiviott 2007; Yusuf 2001). Although clopidogrel remains the most widely used P2Y12 inhibitor, the more potent P2Y12 inhibitors prasugrel and ticagrelor are favoured for DAPT without anticoagulation following PCI in people with a high thrombotic risk (Ibanez 2018; Roffi 2016).

Approximately 5% to 8% of people undergoing PCI have an indication for long‐term oral anticoagulation, most commonly due to atrial fibrillation (Pérez‐Gómez 2004; Rubboli 2007; Wang 2008). CAD shares similar risk factors (diabetes, hypertension, genetics) with atrial fibrillation (Kohli 2014; Neelankavil 2015). As the population ages, the incidence of CAD and atrial fibrillation that require intervention has increased (Cho 2015). DAPT is not sufficient to reduce the risk of ischaemic stroke in atrial fibrillation (Connolly 2006), therefore people with atrial fibrillation and an increased risk of stroke (CHA₂DS₂VASc score of ≥ 1) are treated with oral anticoagulation (Camm 2010; January 2014; Kirchhof 2016). However, triple therapy by combining DAPT with oral anticoagulation significantly increases the risk of bleeding (Paikin 2010).

Balancing the risk of bleeding and thrombosis post‐PCI in people with an indication for anticoagulation is a challenging clinical issue. Questions regarding the efficacy of non‐vitamin K antagonist oral anticoagulants (NOACs) in combination with antiplatelet agents therefore need to be answered.

Description of the intervention

NOACs are a group of oral anticoagulants that inhibit specific coagulation factors: thrombin or activated factor Xa. NOACs have pharmacologic advantages over vitamin K antagonists (VKA), including absence of food interactions, limited drug interactions, and predictable pharmacokinetics. NOACs can therefore be given at fixed dosing without the need for dietary restrictions or routine coagulation monitoring. NOACs have a rapid onset of action within 0.5 to 4 hours, a faster offset of action, and shorter half‐life, therefore, unlike VKA, NOAC therapy can be initiated without temporary concurrent parenteral anticoagulant such as low molecular weight heparin (LMWH) (Bauer 2013; Mekaj 2015; Zirlik 2017).

While VKAs reduce the synthesis of functional vitamin K‐depending clotting factors II, VII, IX, X as well as protein C and protein S, NOACs directly inhibit an activated clotting factor, either FIIa or FXa. Dabigatran is currently the only direct thrombin inhibitor and was the first NOAC, approved in 2010. Factor Xa inhibitors include rivaroxaban, apixaban, and edoxaban (Mekaj 2015). Some NOACs have not been approved due to safety concerns or ineffectiveness, such as the direct thrombin inhibitor ximelagatran and direct factor Xa inhibitors darexaban and letaxaban (Ahrens 2012; Steg 2011; Wallentin 2003).

NOACs are approved for the prevention of venous thromboembolism (VTE) in orthopaedic patients undergoing hip or knee replacement surgery, the treatment of deep vein thrombosis or pulmonary embolism, and for stroke prevention in non‐valvular atrial fibrillation. Rivaroxaban has also been approved for secondary prophylaxis post‐ACS in Europe (Bayer Pharma AG 2016; European Medicines Agency 2013; Kearon 2016; Kirchhof 2016; Konstantinides 2014). VKA remain the medication of choice in people with valvular atrial fibrillation (due to mitral stenosis), mechanical heart valves, VTE associated with antiphospholipid syndrome, and unstable cancer patients. NOACs are not approved for these indications due to insufficient data on their efficacy and safety (Lee 2016).

NOACs are at least as effective as VKA for the prevention of stroke and systemic embolism in people with non‐valvular atrial fibrillation, and they are associated with a significant reduction of intracranial haemorrhage and mortality (Connolly 2009; Giugliano 2013; Granger 2011; Miller 2012; Patel 2011; Ruff 2014).

NOACs are excreted renally, therefore they are contraindicated in renal failure and should be used with dose reduction in people with renal impairment. Dose adjustments are also necessary in people with low or very high weight. NOACs lack reversal agents, therefore management of serious bleeding can be difficult. However, reversal agents are in development, and idarucizumab (a dabigatran antidote) was recently licensed (Glund 2015; Pollack 2015). Other limitations to the use of NOACs include higher cost compared to VKAs and the lack of laboratory monitoring to objectively assess compliance.

The role of plasmatic coagulation in the atherothrombotic clot formation has recently been gaining importance. The thrombotic complications after PCI are not only due to platelet aggregation, but also to plasmatic coagulation with thrombin‐mediated fibrin formation (Merlini 1994). The addition of VKAs (e.g. warfarin) to aspirin to reduce atherothrombotic recurrences after an ACS increases bleeding complications (Andreotti 2006; Rothberg 2005). As a result, studies have tested novel therapies targeting thrombin‐mediated pathways, including NOACs, to prevent thrombotic complications of ACS. Although trials with dabigatran and apixaban failed to show a significant clinical benefit post‐ACS (Alexander 2009; Alexander 2011; Oldgren 2011), rivaroxaban in addition to DAPT has shown a significant reduction of cardiovascular and overall mortality (Gibson 2011; Mega 2012).

The antithrombotic strategy post‐PCI in people with an indication of oral anticoagulation (most commonly atrial fibrillation) poses a dilemma, as they require both antiplatelet and anticoagulation. As a result, recent trials have assessed different antithrombotic treatment regimens and drug combinations to maintain anticoagulation but minimise bleeding complications. One of these strategies is to withdraw aspirin. The combination of oral anticoagulation and clopidogrel was at least equal to or better in efficacy and safety compared to the combination of oral anticoagulation, clopidogrel, and aspirin (Dewilde 2013; Lamberts 2013). Recent randomised controlled trials have evaluated alternative therapies such as NOACs. The PIONEER AF‐PCI study showed that using rivaroxaban with a single P2Y12 antagonist or with DAPT lowers rates of clinically significant bleeding compared with conventional triple therapy with warfarin plus DAPT (Gibson 2016). In addition, dabigatran and a P2Y12 antagonist without aspirin post‐PCI in people with non‐valvular atrial fibrillation is associated with a reduced risk of bleeding versus warfarin and non‐inferior efficiency in preventing thromboembolic events (Cannon 2017). Ongoing randomised controlled trials investigate therapeutic strategies combining apixaban or edoxaban and antiplatelet therapy in comparison to combination therapy with VKA for people with non‐valvular atrial fibrillation undergoing PCI.

To date, combination therapy with a NOAC and an antiplatelet agent has not been addressed in people with VTE, and it is unclear whether data from atrial fibrillation trials can be extrapolated for these patients when they undergo PCI.

How the intervention might work

Thrombotic complications after PCI have been primarily considered a platelet‐mediated process. However, rates of recurrent cardiovascular ischaemic events still remain high despite use of DAPT regimens. This could be due to increased thrombin generation and associated increased clot formation (Merlini 1994; Moon 2017). NOACs selectively target thrombin (IIa) or activated factor X (Xa), which play a role in the coagulation cascade that leads to fibrin formation and resultant thrombosis. NOACs may therefore be beneficial post‐PCI in preventing thrombus formation (Mackman 2008).

Why it is important to do this review

The number of interventional coronary procedures has increased dramatically in recent years (Camm 2010). Current guidelines recommend triple therapy comprising aspirin, clopidogrel, and oral anticoagulation post‐PCI for people with an indication for anticoagulation and high risk for ischaemic complications in order to reduce their risk of stent thrombosis and thrombotic arterial or venous events (Lip 2014; Valgimigli 2017; You 2012). However, this regimen is associated with a high rate of bleeding complications, and the optimal anticoagulation therapy remains uncertain (Paikin 2010). The role of NOACs in these patients is still poorly understood, and treatment decisions must rely on limited evidence.

We aim to review the evidence on the efficacy of NOACs in people undergoing PCI with an indication for anticoagulation. Given the complexity of the condition, and in the absence of randomised controlled trials comparing different types of NOACs against each other, it is essential to carry out a comprehensive and comparative evaluation of all available treatment options within the framework of a network meta‐analysis assessing which treatments, if any, are the most effective and safe.

Network meta‐analysis has many advantages over conventional pairwise meta‐analysis, as the technique allows for the comparison of all interventions, including those for which head‐to‐head comparisons have not been conducted (Jansen 2008; Lu 2004). It also helps in understanding the amount of available evidence for each treatment and treatment comparison (Mills 2013).

This review will include a network meta‐analysis to compare NOACs against each other by including direct and indirect comparisons from multiple treatment arms and across trials (Figure 1). To our knowledge, no other systematic review and network meta‐analysis has assessed the efficacy of NOACs in people undergoing PCI with an indication for anticoagulation. We therefore aim to present the most current and best available evidence to patients, clinicians, policymakers, and researchers.

Figure 1.

Network plot for randomised controlled trials of NOACs in percutaneous coronary intervention. Circles represent the drug as a node in the network; lines represent direct comparisons using randomised controlled trials; dashed lines represent indirect comparisons.

Abbreviations: APX: apixaban; DAB: dabigatran; EDX: edoxaban; NOAC: non‐vitamin K antagonist oral anticoagulant; RVX: rivaroxaban; VKA: vitamin K antagonist

Objectives

To review the evidence from randomised controlled trials assessing the efficacy of non‐vitamin K antagonist oral anticoagulants (NOACs) post‐percutaneous coronary intervention (PCI) in people with an indication for anticoagulation.

Methods

Criteria for considering studies for this review

Types of studies

We will include parallel‐arm randomised controlled trials. We will include studies reported as full text, those published as abstract only, and unpublished data.

We do not anticipate that we will encounter any cluster‐randomised trials. However, if such trials are identified, they will be included.

Types of participants

We will include adult participants (≥ 18 years) who had PCI for ACS or SCAD and who have an indication for NOACs (for the prevention of stroke in non‐valvular atrial fibrillation or VTE).

We will exclude participants with the following comorbidities/characteristics.

• High risk of bleeding or anticoagulation contraindication (e.g. active or recent clinically significant bleeding, thrombocytopenia, prior intracranial haemorrhage)

• Prior stroke/transient ischaemic attack

• Prior stent thrombosis or stent‐within‐a‐stent

• Atrial fibrillation from reversible conditions (e.g. pulmonary embolism, recent surgery, thyroid abnormalities)

• Active malignancy or life expectancy less than one year

In trials with mixed populations, that is trials where some participants meet the inclusion criteria and others do not, we will attempt to include only the eligible participants if this information was reported separately or could be obtained from trial authors. Otherwise, we will include studies with a mixed population if the majority (> 50%) of the participants meet the eligibility criteria.

Types of interventions

We will include trials comparing any type of NOAC (dabigatran, rivaroxaban, apixaban, edoxaban) with placebo, vitamin K antagonists, or a different type of NOAC.

We will include the following co‐interventions if they are not part of the randomised treatment: single or dual antiplatelet therapy.

We will exclude NOACs that are not licensed by the US Food and Drug Administration or European Medicines Agency due to lack of safety or effectiveness (e.g. ximelagatran, darexaban, and letaxaban), as they are not clinically relevant.

Types of outcome measures

Primary outcomes

• Death from cardiovascular causes

• Myocardial infarction

• Stroke

• Major bleeding as defined by the Thrombolysis In Myocardial Infarction (TIMI) criteria (Mehran 2011)

Secondary outcomes

• Death from any cause

• Stent thrombosis

• Any non‐major TIMI bleeding

• Recurrent hospitalisation (at least one at one year)

• Health‐related quality of life (HRQL) assessed using validated instruments (e.g. 36‐Item Short Form Health Survey (SF‐36), EQ‐5D)

We will assess mortality outcomes, myocardial infarction, stroke, stent thrombosis, and bleeding outcomes at longest point of follow‐up for each trial.

Definitions of clinical events (e.g. myocardial infarction and stroke) outcomes will be according to the individual trialists.

Reporting one or more of the outcomes listed here in the trial will not be an inclusion criterion for the review. Where a published report does not appear to report one of these outcomes, we will access the trial protocol and contact the trial authors to ascertain whether the outcomes were measured but not reported. We will include relevant trials that measured these outcomes but did not report the data at all, or reported the data in an unusable format, in the review as part of the narrative.

Search methods for identification of studies

Electronic searches

We will identify trials through systematic searches of the following bibliographic databases.

• Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library

• MEDLINE (Ovid, from 1946 onwards)

• Embase (Ovid, from 1980 onwards)

• Conference Proceedings Citation Index ‐ Science (CPCI‐S) on Web of Science (Thomson Reuters, from 1990 onwards)

We will adapt the preliminary search strategy for MEDLINE (Ovid) (Appendix 1) for use in the other databases. We will apply the Cochrane sensitivity‐maximising randomised controlled trial filter to MEDLINE (Ovid) and adaptations of it to the other databases, except CENTRAL (Lefebvre 2011).

We will also conduct a search of the US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov) and the WHO International Clinical Trials Registry Platform (ICTRP) (apps.who.int/trialsearch/) for ongoing or unpublished trials.

We will search all databases from their inception to the present and impose no restrictions on language of publication or publication status. If we identify any papers in a language not known by the author group, we will seek assistance. This will be acknowledged in the published review.

We will not perform a separate search for adverse effects of NOACs. We will consider adverse effects described in the included studies only.

Searching other resources

We will check reference lists of all included studies and any relevant systematic reviews identified for additional references to trials. We will also examine any relevant retraction statements and errata for included studies.

Data collection and analysis

Selection of studies

Two review authors (SAS, SA) will independently screen the titles and abstracts of all the records identified as a result of the search for potential inclusion and code them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. A third review author (DD) will be asked to arbitrate in case of disagreement. We will retrieve the full‐text reports/publications, and two review authors (SAS, SA) will independently screen the full texts and identify studies for inclusion, and identify and record reasons for exclusion of the ineligible studies. Any disagreements will be resolved through discussion or by consulting a third review author (DD) if necessary. We will identify and exclude duplicates and collate multiple reports of the same study so that each study, rather than each report, is the unit of interest in the review. We will record the selection process in sufficient detail to complete a PRISMA flow diagram and 'Characteristics of excluded studies' table (Liberati 2009).

Data extraction and management

We will use a data collection form for study characteristics and outcome data that has been piloted on at least one study in the review. One review author (SAS) will extract study characteristics from the included studies. We will extract the following study characteristics.

• Methods: study design, total duration of study, number of study centres and location, study setting, and date of study.

• Participants: N randomised, N lost to follow‐up/withdrawn, N analysed, mean age, age range, gender, inclusion criteria, and exclusion criteria.

• Interventions: intervention, comparison, concomitant medications, and excluded medications.

• Outcomes: primary and secondary outcomes specified and collected, and time points reported.

• Notes: funding for trial, and notable conflicts of interest of trial authors.

Two review authors (SAS, SA) will independently extract the outcome data from the included studies. Any disagreements will be resolved by consensus or by involving a third review author (DD) if necessary. One review author (SAS) will transfer the data into the Review Manager 5 file (RevMan 2014). We will double‐check that the data have been entered correctly by comparing the data presented in the systematic review with the data extraction form. A second review author (SA) will spot‐check study characteristics for accuracy against the trial report.

Assessment of risk of bias in included studies

Two review authors (SAS, SA) will independently assess the risk of bias for each study using the criteria outlined in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). Any disagreements will be resolved by discussion or by involving a third review author (DD) if necessary. We will assess risk of bias according to the following domains.

• Random sequence generation

• Allocation concealment

• Blinding of participants and personnel

• Blinding of outcome assessment

• Incomplete outcome data

• Selective outcome reporting

• Other bias (e.g. conflicts of interest or imbalance in baseline characteristics between study arms)

We will grade each potential source of bias as high, low, or unclear and provide a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We will summarise the 'Risk of bias' judgements across different studies for each of the domains listed. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the 'Risk of bias' table. We will group outcomes by method of measurement into objective (e.g. overall mortality), semi‐objective (e.g. cause‐specific mortality, myocardial infarction, and stroke), and subjective outcomes (e.g. HRQL). We will assess attrition bias and blinding of participants, personnel, and outcome assessors for each of these groups separately.

When considering treatment effects, we will take into account the risk of bias for the studies that contribute to that outcome.

Assessment of bias in conducting the systematic review

We will conduct the review according to this published protocol and report any deviations from it in the 'Differences between protocol and review' section of the systematic review.

Measures of treatment effect

We will analyse dichotomous data as risk ratios with 95% confidence intervals (CI) and continuous data as mean differences with 95% CIs (standardised mean differences if data are not available on the same scale). We will enter data presented as a scale with a consistent direction of effect.

Unit of analysis issues

Our unit of analysis will be the participant. For cross‐over trials, we will only include data from the pre‐cross‐over phase. If trials compare more than two intervention groups, we will divide the participants in the control group into two or more groups for the pairwise meta‐analysis; in this way, we will avoid double‐counting participants in the control group, as a serious unit of analysis problem arises if the same group of participants is included twice in the same meta‐analysis. If outcome measures are reported at multiple time points, we will extract all the data reported. We will present the longest point of follow‐up for each trial. If we identify cluster‐randomised trials, we will analyse them according to the methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions and in Cooper 2012 in order to incorporate the results from any cluster‐randomised trials into the network meta‐analysis (Higgins 2017).

Dealing with missing data

We will contact investigators or study sponsors in order to verify key study characteristics and to obtain missing numerical outcome data where possible (e.g. when a study is identified as abstract only). We will calculate missing standard deviations from other statistics such as standard errors, CIs, or P values, according to the methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We will not impute any other missing values.

Assessment of heterogeneity

We will use the I² statistic to measure heterogeneity among the trials in each analysis. If we identify substantial or considerable heterogeneity, we will report it and explore possible causes by prespecified subgroup analysis.

We will follow the recommendations for threshold outlined in Section 9.5.2 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2017; Higgins 2017).

• 0% to 40%: might not be important.

• 30% to 60%: may represent moderate heterogeneity.

• 50% to 90%: may represent substantial heterogeneity.

• 75% to 100%: considerable heterogeneity.

We will investigate possible heterogeneity through subgroup analyses. We will ultimately consider not conducting an overall meta‐analysis if the subgroup analysis shows different effects and the overall meta‐analysis shows substantial or considerable statistical heterogeneity (assessed by visual inspection of forest plots and I² statistic) (Deeks 2017; Higgins 2017).

In network meta‐analyses, inconsistency occurs when direct and indirect estimates do not agree. Heterogeneity and inconsistency will be inspected in the planned network meta‐analyses by decomposing the Q statistics and creation of a net heat plot using the R package netmeta and assessing differences between direct and indirect estimates using descriptive z‐tests. In case of heterogeneity or inconsistency, or both in network meta‐analyses among the primary outcomes, we will explore possible sources and conduct subgroup analyses using determined sources of heterogeneity and/or inconsistency as effect modifiers.

Assessment of reporting biases

If we are able to pool more than 10 trials, we will create and examine a funnel plot to explore possible small‐study biases for the primary outcomes.

Data synthesis

We will undertake meta‐analyses only where this is meaningful, that is if the treatments, participants, and the underlying clinical question are similar enough for pooling to make sense. We will pool the results for HRQL using a mean difference (standardised mean difference if data are not available on the same scale).

We will use a random‐effects model in our meta‐analyses. The random‐effects model is based on the assumption that the true effect might vary from study to study, and the CIs for the average intervention effect will be wider than those obtained using a fixed‐effect approach, leading to a more conservative interpretation.

Network meta‐analysis uses both the direct and indirect evidence in the network to estimate all treatment comparisons in the network even where no head‐to‐head evidence exists. We will conduct a network meta‐analysis based on direct comparisons to generate indirect comparisons of anticoagulants across treatment trials.

We will graphically evaluate the geometry of the whole network to determine whether a network meta‐analysis is feasible. We will also consider whether the trials and populations are sufficiently similar to be included in a network meta‐analysis, so that we can assume consistence in true treatment effects. We will conduct random‐effects network meta‐analyses using the R package netmeta (Rücker 2017). We will present results from network meta‐analyses as summary relative effects (RR) for each possible pair of treatments. In cases of zero events in one or more arms, we will apply the constant continuity correction approach (continuity correction 0.5) and will investigate the influence of continuity correction using further sensitivity analyses (Sweeting 2004). Competing treatments will be ranked by P scores, which allow ranking of treatments on a continuous 0‐to‐1 scale and are derived from the P values of all pairwise comparisons. P scores measure the mean extent of certainty that a treatment is better than the competing treatments. This interpretation is comparable to that of the surface under the cumulative ranking curve (SUCRA) (Rücker 2015).

'Summary of findings' table

We will create a 'Summary of findings' table using the following outcomes: death from cardiovascular causes, myocardial infarction, stroke, major bleeding, death from any cause, and stent thrombosis. We will use the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to assess the quality of a body of evidence as it relates to the studies that contribute data to the meta‐analyses for the prespecified outcomes. We will furthermore apply the four‐step approach presented by Puhan to rate the quality of evidence in each of the direct, indirect, and network meta‐analysis estimates (Puhan 2014).

Overall, we will use methods and recommendations described in Section 8.5 and Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017; Schünemann 2017), employing GRADEpro software (GRADEpro GDT). We will present a 'Summary of findings' table for each comparison (such as NOAC versus vitamin K antagonists or NOAC versus placebo). In the 'Summary of findings' tables we will present direct evidence, indirect evidence, and results from the network meta‐analyses and ranking of treatments using P scores (Brignardello‐Petersen 2017; Brignardello‐Petersen 2018; Puhan 2014).

Two review authors (SAS, SA) will independently make judgements about the quality of the evidence, with disagreements resolved by discussion or by involving a third review author (JM) if necessary. Judgements will be justified, documented, and incorporated into the reporting of results for each outcome.

We plan to extract study data, format our comparisons in data tables, and prepare a 'Summary of findings' table before writing the results and conclusions of our review. A template 'Summary of findings' table for the NOACs versus vitamin K antagonists comparison is presented in Table 1.

Table 1.

'Summary of findings' table

Non‐vitamin K antagonist oral anticoagulants (NOACs) post‐percutaneous coronary intervention in people with an indication for anticoagulation
Patient or population: adults who underwent percutaneous coronary intervention and who have an indication for anticoagulation Settings: secondary care Intervention: NOACs Comparison: vitamin K antagonists
Comparison	Direct evidence			Indirect evidence		Network meta‐analysis
	No. of participants (studies)	OR (95% CI)	Quality of the evidence	OR (95% CI)	Quality of the evidence	OR (95% CI)	Quality of the evidence
1. Death from cardiovascular causes
NOACs vs VKA			◯◯◯◯		⨁◯◯◯◯		⨁◯◯◯◯
DAB vs VKA			⨁◯◯◯◯		⨁◯◯◯◯		⨁◯◯◯◯
RVX vs VKA			⨁◯◯◯◯		⨁◯◯◯◯		⨁◯◯◯◯
APX vs VKA			⨁◯◯◯◯		⨁◯◯◯◯		⨁◯◯◯◯
EDX vs VKA			⨁◯⨁◯◯		⨁◯◯◯◯		⨁◯◯◯◯
2. Myocardial infarction
NOACs vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
DAB vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
RVX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
APX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
EDX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
3. Stroke
NOACs vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
DAB vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
RVX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
APX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
EDX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
4. Major bleeding as defined by the Thrombolysis In Myocardial Infarction (TIMI) criteria
NOACs vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
DAB vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
RVX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
APX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
EDX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
5. Death from any cause
NOACs vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
DAB vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
RVX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
APX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
EDX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
6. Stent thrombosis
NOACs vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
DAB vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
RVX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
APX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯
EDX vs VKA			⨁◯◯◯		⨁◯◯◯		⨁◯◯◯

Abbreviations: APX: apixaban; CI: confidence interval; DAB: dabigatran; EDX: edoxaban; NOAC: non‐vitamin K antagonist oral anticoagulant; OR: odds ratio; RVX: rivaroxaban; VKA: vitamin K antagonist

GRADE Working Group grades of evidence. The quality of the evidence is categorised as high ⊕⊕⊕⊕, moderate ⊕⊕⊕⊝, low ⊕⊕⊝⊝, or very low ⊕⊝⊝⊝. High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.

Subgroup analysis and investigation of heterogeneity

We plan to carry out the following subgroup analyses for the primary outcomes.

• Indication of NOAC: non‐valvular atrial fibrillation and VTE.

• Indication for PCI: STEMI, NSTEMI, unstable and stable angina.

• Risk of stroke: CHA₂DS₂‐VASc score < 2, 2 to 4, > 4.

• Kind of coronary stents: dual therapy stent, bioresorbable vascular scaffold, bio‐engineered stent, drug‐eluting stent, bare metal stent.

• Evaluation of the involved coronary vessel.

• Mean age of the participants in each trial: elderly (≥ 75 years of age) versus non‐elderly (< 75 years of age).

We will use the formal test for subgroup interactions in Review Manager 5 (RevMan 2014).

Sensitivity analysis

We plan to carry out a sensitivity analysis to assess the effect of excluding studies judged as being at unclear or high risk of bias in any of the 'Risk of bias' domains.

Reaching conclusions

We will only base our conclusions on findings from the quantitative or narrative synthesis of the studies included in this review. We will avoid making recommendations for practice, and our implications for research will suggest priorities for future research and outline what the remaining uncertainties are in the area.

Appendices

Appendix 1. Preliminary MEDLINE (Ovid) search strategy

1 Angioplasty, Balloon, Coronary/

2 (balloon adj3 angioplast*).tw.

3 (percutaneous adj3 coronary).tw.

4 PCI.tw.

5 PTCA.tw.

6 (coronary adj5 angioplast*).tw.

7 (coronary adj5 stent*).tw.

8 ((transluminal or trans‐luminal) adj6 coronary).tw.

9 (coronary adj5 balloon dilation*).tw.

10 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9

11 ((novel or new) adj5 anticoagulant*).tw.

12 NOAC*.tw.

13 DOAC*.tw.

14 ((non‐vitamin K or direct) adj5 oral anticoagulant*).tw.

15 apixaban.tw.

16 dabigatran.tw.

17 rivaroxaban.tw.

18 edoxaban.tw.

19 dabigatran/ or rivaroxaban/

20 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19

21 10 and 20

22 randomized controlled trial.pt.

23 controlled clinical trial.pt.

24 randomized.ab.

25 placebo.ab.

26 drug therapy.fs.

27 randomly.ab.

28 trial.ab.

29 groups.ab.

30 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29

31 exp animals/ not humans.sh.

32 30 not 31

33 21 and 32

Contributions of authors

S Al Said conceived, designed, co‐ordinated, and drafted the protocol.

S Alabed and D Duerschmied designed and drafted the protocol.

K Kaier, JJ Meerpohl, and C Bode provided clinical expertise and general advice and revised the protocol.

All authors approved the final version of the protocol and will do the same with the final manuscript.

Sources of support

Internal sources

• No sources of support supplied

External sources

• This project was supported by the National Institute for Health Research (NIHR), via Cochrane Infrastructure funding to the Heart Group. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, National Health Service (NHS), or the Department of Health, UK.

• National Institute for Health Research, UK.

  S Alabed currently holds an NIHR Academic Clinical Fellowship (ACF)

• This project was supported by the Complex Reviews Support Unit, funded by the National Institute for Health Research (project number 14/178/29), UK.

Declarations of interest

S Al Said: none known.

S Alabed: none known.

K Kaier: none known.

C Bode: received research grants from Bayer, GlaxoSmithKline, and Merck; speaker’s honoraria from Bayer, Bristol‐Myers Squibb/Pfizer, Daiichi Sankyo, Boehringer Ingelheim; and consulting fees from Bayer.

JJ Meerpohl: none known.

D Duerschmied: received speaker's honoraria from Bayer Healthcare, Pfizer and Daiichi Sankyo, Travel support for national and international cardiology meetings from Bayer Healthcare and Daiichi Sankyo and support for the organization of local scientific meetings from Bayer Healthcare, Pfizer and Daiichi Sankyo.

New